Tag Archives: free fall

Free Fall, Part 2: Murder on the Zinderneuf

Murder on the Zinderneuf

Mystery stories have been a staple of adventure gaming since 1978’s Mystery Mansion. That’s little surprise; no other form of traditional static literature so obviously sees itself as a form of game between reader and writer, and thus is so obviously amenable to adaptation into other ludic forms. Said adaptations existed well before the computer age, in such forms as the Baffle Books of the 1920s, the Dennis Wheatley Crime Dossiers of the 1930s, and the perennial board game Cluedo (Clue in North America) of 1949.

The early computerized mystery games had the superficial trappings of classic mystery literature but little of the substance. Games like Mystery Mansion and Mystery House were essentially standard Adventure-style treasure hunts, full of mazes and static puzzles, that happened to play out on the stage set of a mystery story. It really wasn’t possible to implement much else with, say, On-Line’s primitive Hi-Res Adventure engine.

That, of course, is why Infocom’s Deadline came as such a revelation. Unlike virtually everyone else making adventure games as of 1982, Infocom had the tools to do a mystery right, to capture the spirit and substance of classic mystery stories in addition to the window dressing. With such a proof of concept to examine (and one which proved to be a major hit at that), combined with a recent uptick in interest in the mystery genre within ludic culture in general following the republication of the old Dennis Wheatley dossiers and an elaborate new board game called Sherlock Holmes: Consulting Detective, other developers started diving into mysteries with similar earnestness. Some of them worked in the text-adventure form, but others branched out into other paradigms. For instance, Spinnaker’s two child-oriented Snooper Troops games and CBS Software’s two adult-oriented Mystery Master games replaced parsers and a single complex story with a more casual form of crime solving. Each contains a series of shorter cases to solve by traveling around a graphical city map, ferreting out clues at each location using a menu-driven interface. A top rating is achieved by solving the crime quickly, using a minimum of clues.

And then there was the game that would become known mostly as that other Free Fall game after the huge success of Archon: Murder on the Zinderneuf. It’s that most interesting anomaly that pops up more than you might expect, an adventure game designed by someone who didn’t much like adventure games.

Jon Freeman laid out his objections to traditional adventure games in an article in the December 1980 issue of Byte, contrasting the form and its limitations with those of the CRPG form he was then working with in crafting Automated Simulations’s DunjonQuest games. An adventure game, he says, is so static that it’s hardly a game at all. It’s “really a puzzle that, once solved, is without further interest.” The former part of this claim became increasingly less true as more dynamic, responsive game worlds like that of Deadline were developed, but the latter part… well, it’s hard to deny that point. The real question is to what extent this bothers you. One remedy to this fundamental failing is perhaps to create longer, deeper works that take as long to play once as it might take you to exhaust the interest of another type of game over many, many plays. Another, of course, is to simply say so what, to note that no one ever criticizes other forms of art, like books, for not being infinitely re-readable (not that Shakespeare doesn’t come close). But still, a re-playable adventure (or for that matter re-readable book) would, all else being held equal, be superior to a non-re-playable version of the same game. Freeman, who still lists Cluedo amongst his favorite games of all time, recycled that game’s concept on the computer, but fleshed out the suspects, the setting, the randomly generated stories behind the murders themselves, to make something more in line with the expectations of adventure gamers.

The mystery may change, but the setting and the actors, the raw materials of these little computer-generated dramas, must inevitably remain the same. Luckily, they’re pretty inspired. The game takes place in 1936, the heyday of the rigid airship, surely one of the most romantic and just plain cool methods of travel ever invented. On a trans-Atlantic voyage aboard the fictional German airship Zinderneuf, a murder has been committed. Which of the sixteen passengers was killed, and which did the killing, and why… these are the elements that are generated anew each time. As a whole genre of pulp-action tabletop RPGs have taught us, the 1930s are a wonderful period for fans of intrigue and derring-do, and Zinderneuf uses that well. Freeman and Reiche work in a lot of the era’s touchstones: old Hollywood, action serials, the Berlin Olympics, the Spanish Civil War, the mob, Amelia Earhart, spiritualism, adventurous archaeologists (Raiders of the Lost Ark was still huge while they worked on the game), and of course Communists and Nazis. It’s an effervescent, pulpy version of history. (That said, our libertarian friend Freeman just can’t restrain himself from taking a political shot at Franklin Delano Roosevelt that strikes a weird sourpuss note amongst all the fun: “Roosevelt was still offering his own version of ‘bread and circuses’ as he ‘guided’ the United States through an unprecedented four terms of depression and war.”) The Zinderneuf itself, meanwhile, proves perfect for a Murder on the Orient Express-style whodunnit. Playing as one of eight detectives drawn from literature or television — including homages to Mike Hammer, Miss Marple, Columbo, and the inevitable Sherlock Holmes among others — you have twelve hours to solve the case before the Zinderneuf touches down in New York and the suspects all scatter to the winds.

Murder on the Zinderneuf

Those twelve hours translate to just 36 minutes of game time — yes, this is a real-time game. The idea here was to replace a 40-hour adventure game with a half-hour game that “can be replayed 100 times.” Also replaced are the text and parser, with a top-down graphical display and an entirely joystick-driven interface.

Murder on the Zinderneuf

Each game begins by telling you who has been murdered from among the cast of characters, each of whom receives a capsule bio in the manual. And then, as Holmes would say (and the manual happily quotes), the game is afoot. You collect evidence in two ways. First, you can search the cabins of the victim and any of the other passengers to see what connections you can discover.

Murder on the Zinderneuf

In the case above, I now know that the murderer of Oswald Stonemann is most likely someone with black hair; the victim is always assumed to have been killed in his cabin. This immediately narrows the suspect list down to five. A logical next step may be to search the cabins of those five suspects, to see what further connections I can turn up. Eventually, however, I will want to start questioning suspects. I can choose the approach I take to each. Various approaches are more or less favorable to different combinations of detective and suspect, something that must be deduced with play. If I choose wisely, perhaps I get a clue.

Murder on the Zinderneuf

Murder on the Zinderneuf

When I believe I have determined opportunity and motive (the game is oddly uninterested in the actual means of murder), I can accuse someone. A false accusation, or one based on insufficient evidence, doesn’t end the game, but does greatly affect your “detective rating” at the end, and prevents you from using that suspect as a source of information for the rest of the game. If you haven’t accused anyone by the time twelve hours (i.e., 36 minutes) have passed, you get one last chance to make an accusation, at some cost to your detective rating, before the game reveals the murderer for you.

There’s much that’s very impressive here. The randomly-generated cases go far beyond the likes of Colonel Mustard in the drawing room with the pistol. Most of the cases don’t even involve that most reliable standby of the mystery writer, love triangles. One time I discovered that Phillip Wollcraft, the archaeologist, had killed the young Natalia Berenski because he was in thrall to certain nameless be-tentacled somethings and needed a handy virgin to sacrifice. (Yes, even the H.P. Lovecraft mythos makes an appearance in this giddy pastiche of a setting, marking what may just be its first appearance in a computer game.) Another time I discovered that the beautiful pilot and all-around adventuress Stephie Hart-Winston had killed the Reverend Jeremiah Folmuth after learning he had in turn killed her beloved brother in a hit-and-run car accident years before. Other cases involve espionage (a natural given the time period), blackmail, even vampires. Most manage to tie the crime back to the period and setting and the specific persona of the characters involved with impressive grace.

But for all that, and despite its superficially easy joystick-driven interface and bright and friendly onscreen graphics that actually look much nicer (at least on the Atari) than those of Archon, Zinderneuf doesn’t quite work for me. Part of the problem derives from all of that rich background information existing only in the manual, not on the screen. The first half-dozen times you play you’re frantically flipping through the pages trying to figure out just who is who as the clock steadily ticks down, an awkward experience a million miles away from Trip Hawkins’s ethos for a new, more casual sort of consumer software. By the time you get over that hump, some of the seams in the narrative generator are already starting to show. You learn what combinations of clues generally lead where, and start to see the same motives repeat themselves. For all the game’s narrative flexibility, there are just eight master stories into which all of the other elements must be slotted. The shock of Wollcraft doing the deed diminishes considerably after you see the same story repeat itself again, with only the name of his victim changed. All of these limitations are of course easily understandable in light of the 48 K of memory the game has at its disposal. Still, things started feeling very shopworn for me long before Freeman’s ideal of a hundred plays.

I also found other elements of the design problematic. When you get down to it, there just isn’t that much to really do, and what there is is often more frustrating than it needs to be. Searching a cabin requires wandering about it trying to cover every square inch until the game beeps to inform you that you discovered a clue — or did not. And talking to suspects can be just as off-putting. Most will only answer a question or two before wandering off again; you then aren’t allowed to speak to them again without speaking to someone else first. Thus the game quickly devolves into a lot of sifting through denials and non-committals, struggling to figure out the right approach to use, while only being able to field one or two questions to your star witness (or suspect) at a time. The memory limitations so strangle the dialog that it’s impossible to pick up clues, as you might in a real conversation, about whether or why your current interrogation approach is failing, or which one might better suit. Murder on the Zinderneuf is fascinating and groundbreaking as a concept, but ultimately a game should be fun in addition to any other virtues it might possess, and here I’m just not sure how well it succeeds. Reading the manual with its cast of exaggerated characters was for me almost more entertaining than actually playing.

Zinderneuf‘s ideal of a narrative that is new every time is neat, and certainly interesting for someone like me to write about as the road almost entirely not taken in adventure games. But are there perhaps good reasons for it to be the road not taken? Maybe for someone primarily interested in games as experiential fictions a 40-hour story, crafted by a person, is more satisfying than 100 30-minute stories generated by the computer. At risk of making Freeman a straw man for my argument, it’s tempting to think again about the flaws that he believed he saw in existing adventures. I believe that designers who see games as rules systems to be carefully crafted and tweaked are often put off by adventure games, which are ultimately all about the fictional context, the lived experience of playing the protagonist in a story. Perhaps having the system itself generate the story could be seen, consciously or unconsciously, as a way to fix this perceived imbalance, to return the art of game design (as opposed to fiction-authoring) to the center of the equation. Yes, Murder on the Zinderneuf‘s narrative generator is clever, but it’s not as clever as, say, Marc Blank, the author of Deadline — and arguably not clever enough to sustain a genre whose appeal is so deeply rooted in its fiction. Zinderneuf is more interesting as a system than as a playable story, in a genre whose appeal is so rooted in story. That, anyway, is how this story lover sees it. Which isn’t to discount Zinderneuf‘s verve in trying something so new. We need our flawed experiments just as much as we do our masterpieces, for they push boundaries and give grist for future designers’ mill. (In that spirit, check out Christopher Huang’s An Act of Murder sometime, which does in text much of what Zinderneuf does in graphics, with results I find more satisfying.)

For several years after 1983, their landmark year of Archon and Murder on the Zinderneuf, Free Fall remained a prominent presence in the growing games industry. In 1984 they released Adept, a sequel to Archon that didn’t quite attract the same love or sales, but was nonetheless a solid success. Soon after they were given an early prototype of the Amiga, thanks to an arrangement Trip Hawkins, a great booster of that machine, worked out with Commodore. Their superb port of Archon became one of the first games available for the Amiga, and they followed it shortly after with a port of Adept of similar quality. Many players still consider these the definitive versions of both games.

Freeman also became a prominent voice in the emerging field of game-design theory, which was separating itself at last by the mid-1980s from the very different art of game programming. He, a defiant non-programmer who had written three books and numerous articles about the art of board-game design before founding Free Fall, was ideally suited to push that process along. Like the last designer I profiled, Dan Bunten, Freeman was given a soapbox of sorts via a column (“The Name of the Game”) in Computer Gaming World. Its ostensible purpose was to tackle tough, controversial subjects head-on. Yet there’s a thin line between delivering hard-hitting, unvarnished reality as one sees it and, well, just kind of sounding like a jerk, and I’m not sure Freeman always stays on the right side of it. His hilarious rant about the Commodore 64 proves that, whatever else he may be, he is no Nostradamus: “software developers will jump off the bandwagon even faster than they got on”; buyers “will think all computers are horrible and throw the whole idea out the window along with their 64.” The Commodore 64 has always evoked special rage from Atari 8-bit loyalists like Freeman. The Atari machines were the 64’s most obvious competitor as fellow low-cost home computers with excellent graphics and sound after weaker sisters like Texas Instruments left the market. They were also arguably the ones the 64 most damaged commercially. “There but for the 64 could have gone the Atari 8-bits,” Atari fans think when they see the 64’s huge success, and not without some justification. But Freeman’s, shall we say, strongly held opinions extended beyond the platform wars. Arcade clones are not just uncreative but morally bankrupt, “illegitimate,” “nasty little pieces of trash.” Programmers doing ports are people “who can’t come up with original subjects for games.” More generally, phrases like “colossal stupidity” and “I almost certainly know more — probably a lot more — about this than you do” creep in a bit too often.

Following the Amiga Archon ports, Free Fall worked for several years on a project that marked a return to Freeman’s roots with Automated Simulations and Temple of Apshai: Swords of Twilight, an ambitious RPG for the Amiga that finally appeared in 1989. It had the unique feature of allowing up to three players to inhabit its world at the same time, each with her own controller, adventuring cooperatively. Despite being released once again by EA, the game seemed to suffer from a dearth of distribution or promotion, and came and went largely without a trace, and without ever being ported beyond the Amiga, a relative minority platform in North America. Another five years elapsed before Free Fall released Archon Ultra, this time on the SSI label. That game was poorly received as adding little to the original, and once again sank quickly into obscurity. And, a few casual card games and the like aside, that’s largely been that from Free Fall. They are still officially a going concern, but seem to exist today largely to license their intellectual property (i.e., Archon) to interested developers. If their output after 1986 or so seems meager given the extraordinary productivity and energy of their first few years, know that my impression — and I must emphasize that this is only an impression, with little data to back it up — is that life has thrown its share of difficulties at Freeman and Westfall since their heydays as stars of Hawkins’s stable of software artists, difficulties that go beyond just some games that performed disappointingly in the marketplace.

If you’d like to try Murder on the Zinderneuf for yourself, I’ve prepared the usual care package for you, with an Atari 8-bit disk image and the (essential) manual. Next time we’ll say goodbye to EA’s Software Artists for a while and catch up with some Implementors again.

(A good interview with Freeman and Westfall can be found online at Halcyon Days, and one with Freeman alone at Now Gamer. Contemporary articles about Free Fall are in the January 1983 Softline, the November 1984 A.N.A.L.O.G., the February 1985 Family Computing, the July/August 1987 Info, and the November 1984 Compute!’s Gazette (Freeman must have been gritting his teeth through that interview, given his opinion of the Commodore 64). Freeman’s Computer Gaming World column ran from the May/June 1983 issue through the April/May 1985 issue.)


Posted by on February 26, 2013 in Digital Antiquaria, Interactive Fiction


Tags: , , ,

Free Fall, Part 1: Archon

Jon Freeman and Anne Westfall

Jon Freeman and Anne Westfall

In the late 1970s Anne Westfall, a mother, housewife, and divorcee in her early thirties, started attending Santa Rosa Junior College. With her children “old enough to take care of themselves,” she was looking for a new direction in her life. She sampled a bit of everything on the college’s menu, but fell in love with computer programming via a course in BASIC. More programming courses followed. She became so good at it so quickly that when some members of the faculty were contacted by a local civil-engineering company that was looking to hire programmers for a new software division they hooked her up with a job. Just like that she had a career; she spent the next two years writing programs for surveyors and subdivision planners on the TRS-80.

At the West Coast Computer Faire of March 1980, fate placed her company’s booth next to that of Automated Simulations of Temple of Apshai and DunjonQuest fame. She got to talking with Automated’s co-founder and primary game designer, Jon Freeman, and a spark both creative and romantic was kindled. Before meeting Freeman computer games had never even occurred to her as an interest, much less a career. She vaguely knew of some housed on some large computer systems to which she had access, and had played Space Invaders a few times at a pizza parlor, but that was about it. Yet Freeman apparently made one hell of an advocate. Not only did she and he become an item, but just five months after meeting her he convinced her to quit her secure job to come program games for Automated Simulations. Soon after they were married.

The marriage has survived to this day, but the new job proved more problematic. Westfall was forced to work as a so-called “maintenance programmer,” tweaking and maintaining the DunjonQuest engine. She also found herself at the epicenter of a power struggle of sorts between Freeman and his founding partner, Jim Connelley. From the time of their first game, Starfleet Orion back in 1978, the two men had fallen into an equitable division of roles. Freeman, who had spent years studying and writing about tabletop-game design but did not program, designed the games; Connelley, a professional programmer for years before Automated’s founding, implemented them. Even as the company grew in the wake of Temple of Apshai‘s success and other designers and programmers came aboard, the basic division of labor remained: Freeman in charge of the creative, Connelley in charge of the technical. From the start Connelley had focused on developing a reusable engine for the DunjonQuest line, written in BASIC for maximum portability and maintainability and capable of running on virtually any computer with at least 16 K of memory. But now, inspired by Westfall’s talent, by newer machines like the Atari 400 and 800, and by newer iterations of the CRPG concept like Ultima and Wizardry, Freeman was getting antsy. Automated’s games were being left behind, he said. He pushed to abandon BASIC and rewrite everything from scratch in assembly language, and to stop targeting a one-size-fits-all lowest-common-denominator machine. Connelley flatly refused, preferring to continue churning out more scenarios using the same old engine. Finally, at the end of 1981, it all devolved into litigation, which ended with Freeman and Westfall, along with other partisans from their camp, walking away. (For what it’s worth, Freeman’s camp ultimately proved to be in the right. Plummeting sales of Automated’s increasingly archaic-looking games forced a major change in direction within a year of the split, including the adoption of the much catchier name Epyx and a new focus on flashy games for next-generation platforms like the Commodore 64. But that’s a story for another time…)

Freeman and Westfall decided to form their own little development group, the cleverly titled Free Fall Associates, to develop games and publish them through others. They would stay small to avoid a repeat of the power struggles at Automated, and write exactly the games they wanted for the platform they wanted: the Atari 800, the most audiovisually advanced 8-bit computer on the market. They would work as partners, as Freeman had in the beginning with Connelley — only now Westfall could assume the programmer’s role. Seeing a divide between slow-paced, ugly, off-putting strategy games and flashier but vapid action fare, they decided to try to make games that slotted in between: fast-paced and aesthetically pleasing but with an element of depth.

Tax Dodge

They took pride in making sure their first game was nothing like those Freeman had designed for Automated Simulations. Tax Dodge was a maze game that took advantage of the Atari’s graphics and sound — but don’t call it a Pac-Man clone or even variant lest Freeman, who railed against the unoriginal arcade clones that still littered the bestseller charts, get very huffy with you. The maze now spanned many screens, smoothly scrolling with the player, an effect that would have been very difficult to manage on the more limited hardware of, say, the Apple II. This gave a quality of exploration, of discovery as the player charted the maze. Rather than ghosts, the player must avoid five sinister IRS agents; rather than gobble pills, she collects cash. Finding an accountant in the maze yields a precious tax shelter. It was a theme near and dear to the heart of Freeman, whose capsule biographies in his games never failed to mention his belief in libertarianism and anarcho-capitalism. Indeed, Freeman was among if not the first designer to sneak political statements into his games. (You may remember his 1980 game Rescue at Rigel, which set players on a hostage-rescue mission against a thinly disguised Ayatollah Khomeini, from an earlier article on this site.)

Tax Dodge made little commercial impression, for which Freeman later blamed the fact that the Atari’s demographics skewed much younger than those of the Apple II and TRS-80, the machines on which Automated had largely concentrated their efforts. Most potential players, he argued, missed the satire that was so much of the fun. Still, it also couldn’t have helped that the game was distributed by a tiny publisher called Island Graphics, who lacked the wherewithal to get the game the sort of prominent advertising and feature reviews that were becoming increasingly important as the software industry steadily professionalized. Maybe this freelance-developer thing wasn’t going to be that easy after all. But then Trip Hawkins and Electronic Arts came calling.

Given that Freeman was one of the few prominent designers not bound by contract to another publisher at the moment, Free Fall was an obvious target for Hawkins in his quest for “software artists.” But they were also a good fit in other ways. If you were reminded of Hawkins’s mantra of “simple, hot, and deep” software when I mentioned Free Fall’s determination to bridge the gap between strategy and action, congratulations, you’ve been paying good attention to my recent articles. Clearly these people were all on the same page. Freeman and Westfall were so excited by Hawkins’s vision that they pitched him two radically different ideas for games. One was for a vaguely chess-like strategy game which would erupt into player-against-player action when two pieces met one another on the board; the other was for an infinitely replayable whodunnit mystery. Hawkins was in turn so impressed that he asked for them both for EA’s stable of launch titles, leaving Free Fall with barely six months to make two ambitious games from scratch.

Freeman and Westfall realized they would need some help. They hired a programmer with whom they had worked at Automated Simulations, Robert Leyland, to implement the mystery, freeing Westfall to just work on the strategy game. And they brought in another person they knew from their Automated days, Paul Reiche III, to work with Freeman on the design of both games.

Reiche was just 22, but had already had quite a career in both tabletop and computer games. While still teenagers, he and some friends had written and self-published a series of supplements for Dungeons and Dragons and other tabletop RPGs. Soon after, TSR themselves came calling, to sweep him off to their Wisconsin headquarters to work for them, doing design, writing, illustrating, whatever was needed. He was undoubtedly talented, but it couldn’t have hurt that, being still a teenager at the time of his hiring, he was willing to work cheap. Regardless, it was a dream job for a young D&D nut; he got to share a byline with Gary Gygax himself on the first Gamma World adventure module while just 20 years old.

Reiche first met Freeman at a D&D convention in 1980, where Freeman was demonstrating the DunjonQuest line in an effort to attract the tabletop RPG crowd to this new computerized variant. The two hit it off, and Reiche soon agreed to design a DunjonQuest scenario for Automated, The Keys of Acheron. Then, around the time of Free Fall’s founding, Reiche got himself fired from TSR, according to his telling for raising a stink about the buying of a Porsche as company car for an executive; maybe working cheap was starting to seem less appetizing. He was back in California, studying geology at Berkeley, when Freeman offered him the chance to get back into game design, this time exclusively on the computerized side. He jumped at the chance. Amongst other advantages, it made good sense from a financial perspective. The tabletop RPG industry was already nearing its historical high-water mark by late 1982, but computer games were just getting started.

I’m going to talk in more detail about Archon, the strategy game, today; next time I’ll talk about Murder on the Zinderneuf, the mystery.

Like so much else, much of the fascination amongst gamers with more, shall we say, colorful variants of chess can be traced back to Star Wars — in this case, to the holographic game played between Chewbacca and R2-D2 aboard the Millennium Falcon. That scene, combined with the explosion in popularity of D&D and by extension fantasy of all stripes, led to a minor craze for new variants of chess. Sometimes that meant nothing more than standard chess sets which replaced pawns with goblins and bishops with dragons to give it all a bit of a different flavor. But other people were more ambitious. The movement reached a sort of absurd fruition when Gary Gygax published the rules for Dragonchess in Dragon magazine’s one-hundredth issue in 1985. It featured a three-level board filled with monsters drawn from D&D‘s Monster Manual, with all of the fiddly rules and exceptions you might expect from the man whose signature game (Advanced Dungeons and Dragons) filled three hardbound rulebooks and hundreds of closely typed pages.

At SCA events and similarly minded gatherings, meanwhile, living chess tournaments became more common. These replaced inanimate chess pieces with real people decked out in appropriate costumes, standing on a board that filled an auditorium floor. When two pieces met in one of these games they battled it out there on the board for the crowd’s delight. Sometimes these battles were purely for show, but in other cases players were assigned roles based on their understood talent at fencing, from pawn to queen and king. In these cases the battles were for real — or as real as fake swords allow. The inevitable result, of course, was a very different sort of game, as suddenly a lucky or dogged pawn, or a tired knight, could alter the balance and ruin the most refined of traditional chess strategies. Freeman participated in such a game as a pawn, experiencing the new spontaneity firsthand. (He acquitted himself well, managing to kill a fellow pawn and then fight a knight to a draw — i.e., a mutual kill.) The experience got him thinking about doing something similar on the computer. It seemed like just the sort of mix of strategy and action Free Fall was after.

Which is not to say that Freeman and Reiche simply recreated the living-chess experience on the computer. If anything Archon‘s relationship to chess is rather overblown, for Archon is both simpler and more complicated. Movement falls into the former category. Every piece has a maximum number of squares it can move in a turn, and either moves on the ground (meaning it can move only horizontally or vertically and cannot jump pieces) or in the air (meaning it can also move diagonally, and can jump pieces). There is nothing like the complications of, say, the knight in traditional chess. On the other hand, there are more pieces to deal with in Archon, and more places to put them. The board is now 9 X 9 rather than 8 X 8, with the requisite additional two units per side. The larger size was chosen because it fit most neatly on the screen, provided the optimum balance between visibility and strategic possibility, and allowed for three power points to be neatly spaced across the middle of the board. Controlling these three spaces, plus the additional power point located at each edge of the board, wins the game instantly. Alternately, if less strategically, one can win by simply killing all of the opposing player’s units.

The Archon game board. Note the three power points running down the center. Two more are hidden under the wizard and sorceress on the center-left and center-right squares.

The Archon game board. Note the three power points running down the center. Two more are hidden under the wizard and sorceress on the center-left and center-right squares.

The two opposing forces are no longer mirror images of each other. The game is subtitled The Light and the Dark; the Light side (presumably good) has different units with different combat abilities from the Dark side (presumably evil). Some units use a melee attack; others shoot missiles or fireballs; still others, like the banshee, have an area attack that spreads outward from their person; each side has one unit (the wizard or the sorceress) who can cast a handful of spells once each per game.

The board at the dark extreme of the luminosity cycle. Note the contrast with the picture above, which shows the cycle at its mid-point.

The board at the dark extreme of the luminosity cycle. Note the contrast with the picture above, which shows the cycle at its mid-point.

Of the squares on the board, 25 are always light, 25 always dark. However, the remaining 31, including the three central power points, constantly cycle from light to dark and back. This fact is critical to strategy, because light units gain a big advantage when fighting on light squares, and vice versa. Thus the wise player plans her attacks and retreats, her feints and thrusts, around the ever-changing board. Accidentally leaving a powerful piece exposed on the wrong color of square can lead to the worst sort of self-recrimination when your opponent pounces to take out your golem with her goblin. And yes, just as in the live chess match that inspired Freeman, double kills are possible.

A phoenix (Light) and dragon (Dark) battle. Because this fight is taking placing on a light square, the phoenix has a huge advantage; note the difference in the life bars at the edge of the screen.

A phoenix (Light) and basilisk (Dark) battle. Because this fight is taking placing on a light square, the phoenix has a huge advantage; note the difference in the life bars at the edge of the screen.

Still other elements of Archon would never have been possible on the tabletop. For instance, the health of each unit is tracked even outside of the combat screen. It takes a few turns to fully recover from a hard fight, meaning a stubborn opponent can kill your wizard just by throwing enough cannon fodder — i.e., goblins — at it turn after turn. The game clearly wants to be played more quickly, more urgently, even (dare I say it?) less strategically than a classic chess match. You find yourself tossing your units into the fray, not pausing to study every option and plan your next several turns in advance. What with the fast pace and the role that reflexes play, playing Archon with another human feels like really going at it, with little of the cool cerebral feel of chess. I have to believe this is intentional, and certainly it’s a more than valid design choice. Indeed, it’s the prime source of Archon‘s appeal in contrast to a game like chess.

That said, there’s one flaw in the strategic game that bothers me enough to really impact my appreciation for the game as a whole. When playing a relatively close game, it’s all too easy to find yourself in an ugly stalemate, in which each player has just a few units left and neither has any incentive to risk any of them by moving them off of favorably colored squares. At this point both sides are just stuck, until someone loses patience at last and attacks the enemy on one of her favorable squares in the face of long odds indeed, all but guaranteeing sacrificing that piece — and, eventually, losing the game — for the sake of just ending the damn thing already. I’m not sure I have any brilliant suggestion of how this could have been fixed — maybe begin to cycle more squares from light to dark as the number of pieces on the board is reduced, thus forcing more dynamism into the game?; maybe add conditions for a chess-style draw? — but I do know that it needed to have been for me to raise my judgment of Archon from “just” a fun and creative effort to the timeless classic many would have me label it. (Then again, it’s possible I’m just missing something strategically obvious. If you have a solution to this dilemma, by all means tell me about it.)

As you might imagine given the time constraints, Westfall, Freeman, and Reiche worked like dogs on Archon even as Freeman and Reiche also labored over Murder on the Zinderneuf. Free Fall had no offices; everything was done out of Freeman and Westfall’s home in Portola Valley, California. Westfall:

We had a tough schedule at first. For six months we didn’t even read a book or go to the movies, and that’s disaster in our house. We basically worked all the time. At meals we were always discussing the games. How to do this, and what to do about that. We worked from the time we got up until all hours of the night. Then we’d get up the next day, grab a cup of coffee, and go back to work.

Archon had been envisioned from the beginning as a two-player game. However, just a month before they had to turn over the game, EA begged them to add a single-player option, thus saddling Free Fall with the task of coding a complete AI, in addition to everything else that still had to be done, in one month. With so little time and eager to preserve the game’s fast-paced character, they focused on making an AI that was “fast and decent” rather than “slow and perfect.” As Ozark Softscape did for M.U.L.E., they also made it possible for the AI to play itself, a godsend for shop display windows. And then they added one additional groundbreaking feature that has been little remarked since the game’s release. Freeman:

There’s a built-in, self-adjusting difficulty factor in Archon so that if the computer keeps beating up on you, it will get easier and easier. But most people don’t know that because it goes in little tiny increments. By the time it really starts kicking in, players think, “Oh, I’m just getting better.” Well, they are, partly; but partly it’s because the computer is not being as good. But nobody knows that’s there. It’s not something we advertise, but we were aware of the problem.

Just like chess: how do two unequal players play chess? Well, not very well. And there’s not really a great deal you can do about it. If you start taking pieces away, you change the game so radically that you’re not playing chess anymore. Archon is the same way. So we said, we want to do a game in which we can do that without screwing it up.

This very likely marks the first example of adaptive AI in the commercial game industry, a radical step in the direction of friendlier, more accessible gameplay — and in the direction of Trip Hawkins’s vision of consumer software — that deserves to be celebrated more than it has been. It also kind of leaves you wondering whether any victory over the computer was truly earned, a dilemma familiar to many modern gamers. Ah, well… groundbreaking as Archon‘s adaptive AI is, the game is still best experienced with two players, where it all becomes moot anyway.

Released in a striking monochrome sleeve that beautifully presented the theme of Light and Dark, Archon struck a major chord with the public. It became the second-best-selling of those seven EA launch titles, behind only Pinball Construction Set. I strongly encourage you to play it, but I’m not going to provide a download here. Free Fall, you see, is still around as at least a semi-going concern and still licensing variants and remakes, and I don’t want to step on any toes. I’m sure you can find the original game on your own if you’re so inclined. The Atari 8-bit incarnation was the first developed and is thus the best reflection of the original vision for the game, although the Commodore 64 port does look nicer. If you do snag one of these versions from somewhere else, maybe think about buying the latest licensed incarnation as well, if for nothing else than to show your appreciation to Freeman and Westfall.

The other Free Fall game amongst those early titles, Murder on the Zinderneuf, didn’t attract anywhere near as much attention as Archon. Yet in its own way it’s every bit as interesting — perhaps even more so if, like me, you like a strong dose of story in your games. We’ll talk about that game, and wrap up the story of Free Fall, next time.

(I’ll include the main sources I used for researching Free Fall in the concluding article.)


Posted by on February 20, 2013 in Digital Antiquaria, Interactive Fiction


Tags: , , ,